A good subcontractor estimating workflow is one repeatable path from bid invitation to closed job: log the invite, decide to bid, take off the plans, price with your own labor numbers, level the quotes, hand the job to the field, then compare actual hours to the estimate. Most subs break it in the same two places: at the start, where invites get lost, and at the end, where field data never gets back to the estimator.
This guide walks through each step for electrical, plumbing, HVAC/mechanical, masonry, drywall, and other specialty trades. It is written for owners, estimators, and project managers at shops with 1 to 50 people.
TL;DR
- Build the workflow around one record per bid. Invite, plans, takeoff, quotes, price, and outcome all live in one place, not in six folders and a workbook.
- Decide bid or no-bid before you measure anything. Estimating time is your scarcest resource.
- Use the same cost codes in the estimate, the job budget, and the timecards. That is what makes the next bid better.
- Treat AI or automated takeoff as a draft. Check area quantities by hand and trust it more on counts.
- Get your foreman to look at the bid before you submit it, not after you win it.
Table of Contents
- Why a workflow beats a better spreadsheet
- Step 1: Capture every invite in one place
- Step 2: Make the bid or no-bid call first
- Step 3: Read the specs before you measure
- Step 4: Take off the plans in a consistent order
- Step 5: Price with your own labor, not a book number
- Step 6: Level the quotes and close scope gaps
- Step 7: Hand off to the field and close the loop
- Where the workflow usually breaks
- Our perspective: the workflow is not an estimating problem
- Putting it in one system
- FAQ
- Sources
Why a workflow beats a better spreadsheet
Most estimators we talk to have a workbook they built over years. It works. The problem is everything around it: the invite sits in an email thread, the plans sit in a download folder, the vendor quotes sit in a PDF, and the actual labor hours sit in a foreman's truck.
That gap has a cost. A CFMA analysis of more than 500 closed projects found subcontractor profitability swinging from -18% to 90% on actual results, against planned ranges of 0% to 56%. A spread that wide means the estimate is not predicting the outcome. Fixing that starts with the process, not the math.
A workflow is simply the order of work, who touches each step, and what gets recorded. Here is the version we think works for specialty trades.
Step 1: Capture every invite in one place
You cannot bid what you did not see. Invites arrive by email, by planroom portal, by text from a GC you have worked with for ten years, and by phone. Due dates live in the subject line or buried on page three of an attachment.
What to record for every invite:
- GC name and the estimator's contact
- Project name, location, and bid due date and time (with time zone)
- Your trade's scope sections and any addenda received
- Status: new, reviewing, bidding, submitted, won, lost, no-bid
The fix is a single bid log with a due-date view sorted soonest first. A shared inbox plus a spreadsheet will do it if everyone is disciplined. If you want it automated, our write-up on bid invitation tracking for trade subs covers how to stop losing 10 to 15 hours a week to inbox hunting.
Check addenda every time. A missed addendum on a glazing or fire protection package can change the scope after you have priced it.
Step 2: Make the bid or no-bid call first
Every takeoff you start costs hours. Decide before you spend them.
A quick scorecard keeps the decision honest. Give each item a 1 to 5:
- Scope fit. Is this work your crews do well, or are you stretching?
- GC relationship. Do they pay on time? Do they hold your numbers or shop them?
- Schedule. Can you man it in the window they want?
- Plan quality. Complete drawings and specs, or "design intent" and a lot of risk?
- Competition. How many subs in your trade are on the list?
Anything under about 15 out of 25 is a no-bid or a high-margin number. Write the reason in the log. After six months you will see which GCs and project types waste your time.
Step 3: Read the specs before you measure
The drawings tell you how much. The spec book tells you what kind, and that is where the money hides.
For your trade, scan these before the first click:
- Division 26 (electrical): conduit types, fixture allowances, lighting controls, testing requirements
- Division 22 and 23 (plumbing, HVAC): pipe material, insulation, balancing and commissioning
- Division 04 (masonry): unit type, mortar, reinforcing, cleaning, mock-up panels
- Division 09 (drywall, painting, flooring): finish levels, fire-rated assemblies, coats, prep
- Division 07 and 08 (roofing, insulation, glazing): warranty terms, system manufacturers, performance ratings
- Division 21 (fire protection): hazard classifications, hydraulic calcs, delegated design
Also read Division 01. Liquidated damages, retainage, prevailing wage, and working-hour limits all change your labor number. Make a short list of exclusions and clarifications as you read. It becomes your proposal's qualifications page.
Step 4: Take off the plans in a consistent order
Takeoff is where speed and accuracy fight each other. A few habits settle it.
Measure in the same order every time. Floor by floor, or system by system, but pick one. Estimators who change the order mid-job double-count.
Set scale on every sheet. A PDF printed at the wrong size is the quiet killer. Our post on scaling PDF drawings locally to prevent doubled quantities shows the common failure.
Use named conditions, not loose measurements. "5/8 Type X, 9 ft ceiling" is a condition. A loose 4,200 SF is a number you will not trust next month. For trade-specific walkthroughs, see our Bluebeam drywall takeoff method and plumbing takeoff from PDF.
Use automation as a first pass. Research on manual versus model-based takeoff in installation projects found that manual, drawing-based takeoff tended to underestimate costs by 8% through human error, while the model-based method was about 40 times more efficient on takeoff time. That study used building models, not a typical PDF set, so do not expect those exact numbers from a scanned plan. The direction is still useful: tools beat tired eyes on counting.
Be careful with AI measuring. A case study on automated quantity takeoff points to a pattern worth knowing: results tend to be strong on counts (outlets, heads, fixtures, doors) and weaker on irregular area-based quantities. So let software count, then verify your big areas and long runs by hand. Treat the output as a draft.
A quick takeoff check
Before you move on, run three sanity checks:
- Does the total area match the gross building area on the plans, within a reasonable margin?
- Does your per-SF count look right against your last three similar jobs?
- Did someone other than you glance at the biggest line item?
Step 5: Price with your own labor, not a book number
Material pricing is a phone call or a supplier quote. Labor is where your bid is won or lost.
Build the labor price from three pieces:
- Production rates from your own completed jobs (feet of conduit per man-day, SF of board hung and finished per crew-day, block laid per mason per day)
- Labor burden: payroll taxes, workers' comp, insurance, benefits, and union or prevailing wage requirements where they apply
- Site factors: height, access, phasing, overtime, and travel
Burden is the line most often wrong. Construction accounting specialists stress that labor is often the largest and least forgiving expense and that estimating should connect to your accounting system so labor burden and hours are tracked against the estimate as the job runs, not reconciled months later. If you use QuickBooks, our guide to QuickBooks for subcontractors and job cost accuracy covers three fixes that keep the numbers matching.
Specific trade workflows go deeper than we can here. See our four-step drywall estimating workflow and the HVAC estimating workflow if either is your trade.
A worked example
Say a drywall sub measures 18,000 SF of board on a three-story office fit-out. The estimator's past jobs show a crew of four hanging and finishing about 1,200 SF per crew-day in similar space. That is 15 crew-days. The estimator then adds 10% for the tight phasing the specs require, bringing it to 16.5 crew-days, and prices burden from the payroll rate sheet. Every number in that chain traces back to a prior job or a spec line. That is the point. When the bid is lost or won, you know why.
Step 6: Level the quotes and close scope gaps
If you buy material, rent equipment, or sub out parts of your scope (a glazing sub hiring a hardware installer, a fire protection sub buying design), you are leveling quotes too.
Compare quotes on scope, not just the bottom line:
- What is included and excluded, line by line
- Delivery, freight, tax, and restocking terms
- Lead times against the schedule
- How long the price is held
A low number that leaves out a major item is not low. Our piece on bid leveling for trade subs explains how scope gaps turn into five- and six-figure losses.
Then build the proposal: base bid, alternates, unit prices, clarifications, and exclusions. Keep a copy of the exact version you sent. When the GC calls three weeks later wanting to "clarify a few things," you need it.
Step 7: Hand off to the field and close the loop
Winning the bid is not the end of the estimate. It is where the estimate gets tested.
The handoff meeting
Before the job starts, sit down with the project manager and the foreman for 30 minutes. Walk through:
- The quantities and where the risk is
- Production rates you assumed
- Exclusions the GC may push back on
- Long-lead items and submittals
- Budget hours by cost code
Industry experts argue the old pattern of estimators building a bid and passing it on with little context is a major cause of failed projects, and that foremen and project managers should weigh in on constructability and site risk during bidding, not after award. If you need a model, this subcontractor selection guide from a Mid-Atlantic general contractor shows how involving field and project teams in preconstruction improves assumptions and handoffs. Even a 15-minute call with your foreman before you submit catches things the plans do not show: a lift that will not reach, a laydown area that does not exist.
Use one set of cost codes
This is the single most useful habit in the workflow. When estimate lines, the job budget, and the foreman's timecards all use the same cost codes, you can compare them directly. The CFMA labor article describes this as bid-to-job continuity, and it addresses a real gap: a 2005 study cited there found roughly 48% of estimators do not routinely feed historical site feedback into estimates.
If your estimate says "rough-in" and your timecards say "labor," you have a black hole. If you want the accounting side explained, read job costing explained for trade contractors, or the electrical job costing mistakes to avoid.
Compare actual to estimate at 25%, 50%, and 75%
Do not wait for job close. At each milestone, pull hours spent against hours budgeted by cost code. If rough-in is at 60% of hours and 40% complete, you know now, while you can still adjust crew size or file a change order. Then, when the job closes, write the real production rate back to your estimating database. That is how the next bid gets better.
Where the workflow usually breaks
| Break point | What it looks like | Fix |
|---|---|---|
| Invites | Missed due dates, duplicate bids | One bid log, due-date view |
| Takeoff | Doubled or missed areas, wrong scale | Named conditions, scale check per sheet |
| Labor pricing | Book rates, wrong burden | Your own production rates, payroll burden |
| Quotes | Low number with a scope gap | Side-by-side leveling |
| Handoff | Foreman learns the plan on day one | Handoff meeting, pre-submit review |
| Feedback | Estimates never improve | Same cost codes, milestone reviews |
Our perspective: the workflow is not an estimating problem
We built Subascent because we kept seeing the same thing in specialty trade shops. The estimating tool was fine. What was missing was continuity. The bid lived in one place, the job in another, and the field hours in a third. By the time the numbers came together, the job was over.
So we think of the estimate as the first draft of the job budget, not a separate document. The same cost structure should run from the bid, to the budget, to the foreman's phone, to the profit report. That is why we focus on subs and not general contractors. A GC's software is built to coordinate many trades on one project. A sub needs the opposite: many projects, one trade, and a tight loop between the office and the crew.
If you do only one thing after reading this, make your estimate codes match your timecard codes. It costs nothing and pays back on the next bid.
— Dave
Putting it in one system
You can run this workflow with a bid log, a PDF viewer, a workbook, and discipline. Many shops do. If the handoffs between those tools are where you lose time, Subascent combines bid tracking, estimating, PDF takeoff (a paid add-on), submittals and RFIs, change orders, daily reports, crew tracking through the free CrewTrack mobile app, QuickBooks sync, and profit analytics for specialty trade subcontractors. See how it works at subascent.com.
FAQ
How long should a subcontractor estimate take?
It depends on scope and plan quality, so there is no honest single number. A better measure is your own: track hours per bid by trade and project type for a month. You will see which bids eat time and which GCs send you work not worth pricing.
Should small subs use software or stick with spreadsheets?
A spreadsheet is fine for pricing logic. The trouble starts when it also has to be your bid log, takeoff record, and job budget. If you are missing due dates or rebuilding the workbook for every bid, that is the sign to move those parts into a system.
What is the most common estimating mistake for specialty trades?
Wrong labor numbers. Book production rates and an incomplete burden figure compound across the whole bid. Use your own completed-job data, and check burden with whoever runs your payroll.
How do I get foremen involved without slowing the bid down?
Send them a one-page summary of the scope, key quantities, and assumptions before you submit. Ask for two things: what looks wrong, and what the plans do not show. Fifteen minutes is usually enough.
Sources
- Labor Overruns: How to Measure & Manage for Peak Performance, CFMA
- Total BIM in Swedish installation contracting, Chalmers University of Technology
- Methodological Case Study in Commercial Construction, EasyChair
- Construction accounting, Fintruction
- Construction Contract Modifications, AGC and USACE joint study